Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A computer-implemented method for simulating occupant behavior for a target building structure, the method comprising: receiving, via a simulation engine, recorded occupant behavior data associated with a representative building structure; constructing a histogram of at least two dimensions that is populated with the recorded occupant behavior data; smoothing the recorded occupant behavior data to fully populate the histogram, wherein the smoothing comprises: determining a smoothing parameter value for each of one or more factors that influences the one or more attributes, and computing a smoothing coefficient for each of the plurality of bins based on the smoothing parameter values and a number of neighboring bins; computing normalized arrays of occupant behavior feature values based on the histogram; generating one or more attributes of occupant behavior for a simulated occupant of the target building structure based on the normalized arrays of occupant behavior feature values; and generating a simulation of occupant behavior based on the generated one or more attributes of occupant behavior, wherein the simulation indicates a predicted energy demand of the target building structure.
This invention relates to simulating occupant behavior in buildings to predict energy demand. The method addresses the challenge of accurately modeling how occupants interact with building systems, which is critical for energy efficiency and demand forecasting. The approach uses recorded occupant behavior data from a representative building to generate a simulation for a target building. The method begins by receiving recorded occupant behavior data, which includes interactions such as lighting, HVAC adjustments, and appliance usage. A multi-dimensional histogram is constructed and populated with this data. To handle sparse data, the method applies smoothing by determining smoothing parameter values for influencing factors and computing smoothing coefficients for histogram bins based on neighboring bins. This ensures the histogram is fully populated. Normalized arrays of occupant behavior feature values are then computed from the smoothed histogram. These arrays are used to generate attributes of simulated occupant behavior, such as usage patterns and timing. Finally, a simulation of occupant behavior is generated, which predicts the energy demand of the target building based on the simulated attributes. The method improves energy demand forecasting by accounting for realistic occupant behavior variations.
2. The method of claim 1 , wherein the one or more attributes include a task performed by the simulated occupant of the target building structure.
This invention relates to simulating occupant behavior in building structures to optimize energy efficiency, safety, or other performance metrics. The problem addressed is the lack of accurate modeling of how occupants interact with building systems, leading to inefficient energy use or suboptimal building design. The invention provides a method for simulating occupants with realistic attributes, including tasks they perform within the building. These tasks influence how the simulated occupants interact with building systems, such as lighting, HVAC, or security systems. By modeling these behaviors, the system can predict energy consumption, identify inefficiencies, or test building designs under realistic conditions. The simulation accounts for variations in occupant tasks, such as working at a desk, moving between rooms, or adjusting environmental controls, to improve the accuracy of performance predictions. This approach enables building designers and operators to make data-driven decisions for optimizing energy use, occupant comfort, and safety. The method can be applied to residential, commercial, or industrial buildings to enhance overall building performance.
3. The method of claim 1 , wherein the one or more attributes include a time duration during which a task is performed by the simulated occupant of the target building structure.
This invention relates to simulating occupant behavior in building structures to optimize energy efficiency, comfort, or other performance metrics. The problem addressed is the lack of accurate modeling of how occupants interact with building systems over time, leading to suboptimal building management. The invention provides a method for simulating occupant behavior by tracking one or more attributes, including the time duration during which a task is performed by a simulated occupant. The simulation models how long occupants engage in activities such as adjusting thermostats, operating lights, or using appliances, allowing for more realistic predictions of energy consumption and comfort conditions. By incorporating time-based attributes, the simulation can better reflect real-world behavior patterns, improving the accuracy of building performance assessments. The method may also include other attributes like frequency, sequence, or environmental triggers to enhance the simulation's realism. The simulated occupant's actions are used to evaluate building system performance, identify inefficiencies, and optimize control strategies. This approach enables building designers and operators to make data-driven decisions for improving energy efficiency and occupant comfort.
4. The method of claim 1 , wherein the one or more attributes include a number of participants associated with a simulated task performed by the simulated occupant of the target building structure.
This invention relates to simulating occupant behavior in building structures to optimize energy efficiency, safety, or other performance metrics. The problem addressed is the need for accurate modeling of how occupants interact with a building, including their activities, movements, and decision-making, to improve building design and operation. The invention involves simulating occupants with realistic attributes to predict their behavior in a target building structure. The method includes simulating one or more occupants within a building structure, where each simulated occupant has attributes that influence their behavior. These attributes include the number of participants involved in a simulated task performed by the occupant. For example, if the task is a group activity, the number of participants affects how the occupant interacts with the building environment, such as energy usage, movement patterns, or safety compliance. The simulation accounts for how different participant counts influence task execution, allowing for more accurate predictions of building performance under various occupancy scenarios. This helps in designing buildings that are more energy-efficient, safer, and better adapted to real-world usage. The simulation may also include other attributes like occupant roles, preferences, or environmental conditions to further refine behavior modeling. The output can be used to optimize building systems, such as HVAC, lighting, or security, based on predicted occupant behavior.
5. The method of claim 1 , wherein generating one or more attributes comprises: computing a probability table based on the normalized arrays of occupant behavior feature values; computing one or more gamma distribution parameters based on the normalized arrays of occupant behavior feature values; and generating a first occupant behavior attribute for a sub interval based on the probability table or the gamma distribution parameters.
This invention relates to analyzing occupant behavior in a monitored environment, such as a building or vehicle, to detect anomalies or optimize resource usage. The method involves collecting sensor data to capture occupant behavior patterns, such as movement, presence, or activity levels, and processing this data to generate behavioral attributes for analysis. The method first normalizes arrays of occupant behavior feature values, which represent quantifiable aspects of behavior derived from sensor inputs. These normalized values are then used to compute a probability table, which statistically models the likelihood of different behavior patterns. Additionally, gamma distribution parameters are calculated to further characterize the behavior data, providing a probabilistic framework for identifying deviations or trends. A key aspect of the invention is generating a first occupant behavior attribute for a specific sub-interval of time. This attribute is derived from either the probability table or the gamma distribution parameters, allowing for detailed temporal analysis of behavior. The method enables real-time or retrospective assessment of occupant behavior, supporting applications such as energy efficiency optimization, security monitoring, or personalized service adjustments. The approach leverages statistical modeling to enhance the accuracy and interpretability of behavior analytics.
6. The method of claim 1 , wherein a number of dimensions of the histogram equals a number of factors that influence the one or more attributes.
This invention relates to data analysis, specifically improving the accuracy of attribute analysis by aligning histogram dimensions with influencing factors. The method involves generating a histogram where the number of dimensions corresponds to the number of factors affecting one or more attributes of interest. By structuring the histogram in this way, the method ensures that each dimension represents a distinct influencing factor, allowing for more precise analysis of how these factors interact to determine attribute values. The histogram may be used to visualize relationships between factors and attributes, identify patterns, or support decision-making processes. This approach enhances traditional histogram-based analysis by explicitly linking dimensions to known influencing factors, reducing ambiguity and improving interpretability. The method is applicable in fields such as data science, machine learning, and quality control, where understanding the impact of multiple factors on attributes is critical. By aligning histogram dimensions with influencing factors, the invention provides a clearer, more structured way to analyze complex datasets.
7. The method of claim 6 , wherein the factors comprise at least one of time of day, the previous task performed by an actual occupant, time elapsed since a task was performed by the actual occupant, and a number of occupants participating in a task.
This invention relates to systems for predicting and managing occupant activities in a smart environment, such as a smart home or office. The problem addressed is the need for intelligent automation that adapts to occupant behavior, improving efficiency and personalization. The invention involves a method for determining the likelihood of an occupant performing a specific task based on contextual factors. These factors include the time of day, the previous task performed by the occupant, the time elapsed since the last task was performed, and the number of occupants participating in a task. By analyzing these factors, the system can predict occupant behavior and trigger automated actions, such as adjusting environmental settings or activating devices, to enhance convenience and energy efficiency. The method may also involve comparing the predicted task with the actual task performed to refine future predictions. This approach enables the system to learn and adapt over time, improving accuracy in anticipating occupant needs. The invention aims to create a more responsive and intuitive smart environment that aligns with occupant routines and preferences.
8. The method of claim 1 , wherein computing the normalized arrays further comprises: calculating a probability value of the occupant performing another task based, at least in part, on performing a previous task that is performed before the another task, wherein the another task, a next task, and the previous task each comprise different tasks and the probability value is calculated based, at least in part, on the recorded occupant behavior data.
This invention relates to systems for analyzing occupant behavior in a vehicle to predict and assist with task performance. The problem addressed is the need to accurately determine an occupant's likelihood of performing a subsequent task based on their prior actions, improving task assistance and automation. The method involves computing normalized arrays of occupant behavior data, which includes calculating a probability value for an occupant performing a future task. This probability is derived from the occupant's recorded behavior data, specifically their performance of a preceding task. The tasks involved are distinct from one another, ensuring that the prediction is based on meaningful behavioral patterns rather than repetitive actions. By analyzing the sequence of tasks—previous, current, and next—the system can anticipate the occupant's needs and provide timely assistance, such as adjusting vehicle settings or suggesting actions. The approach leverages historical data to refine predictions, enhancing the system's ability to adapt to individual occupant behaviors over time. This improves user experience by reducing manual inputs and increasing automation efficiency.
9. The method of claim 1 , wherein computing the normalized arrays further comprises: calculating a probability value of the occupant performing a next task based, at least in part, on performing a previous task that is performed before the next task, wherein the next task and the previous task comprise different tasks and the probability value is calculated based, at least in part, on the recorded occupant behavior data.
This invention relates to systems for analyzing occupant behavior in a vehicle to predict and assist with task performance. The problem addressed is the need to accurately predict an occupant's next task based on their past behavior, improving vehicle automation and user assistance. The method involves computing normalized arrays of occupant behavior data, which includes calculating a probability value for the occupant performing a next task. This probability is determined based on the occupant having performed a previous task, where the previous and next tasks are distinct. The calculation relies on recorded occupant behavior data, which may include historical interactions with vehicle systems, movement patterns, or other behavioral metrics. By analyzing these patterns, the system can anticipate the occupant's likely actions, enabling proactive assistance or automation. The method may also involve preprocessing the recorded behavior data to extract relevant features, such as task sequences or timing, and normalizing these features to ensure consistent analysis. The probability calculation may use statistical models, machine learning, or other predictive techniques to assess the likelihood of task transitions. The system can then use this information to adjust vehicle settings, provide reminders, or automate certain functions, enhancing convenience and safety. The approach improves upon prior systems by incorporating task dependencies and historical behavior to refine predictions.
10. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause a computer system to perform an operation for simulating occupant behavior for a target building structure, the operation comprising: receiving recorded occupant behavior data associated with a representative building structure; constructing a histogram of at least two dimensions that is populated with the recorded occupant behavior data; smoothing the recorded occupant behavior data to fully populate the histogram, wherein the smoothing comprises: determining a smoothing parameter value for each of one or more factors that influences the one or more attributes, and computing a smoothing coefficient for each of the plurality of bins based on the smoothing parameter values and a number of neighboring bins; computing normalized arrays of occupant behavior feature values based on the histogram; generating one or more attributes of occupant behavior for a simulated occupant of the target building structure based on the normalized arrays of occupant behavior feature values; and generating a simulation of occupant behavior based on the generated one or more attributes of occupant behavior, wherein the simulation indicates a predicted energy demand of the target building structure.
This invention relates to simulating occupant behavior in building structures to predict energy demand. The problem addressed is the need for accurate energy consumption modeling, which requires understanding how occupants interact with building systems. Existing methods often lack detailed behavioral data or rely on oversimplified assumptions. The system processes recorded occupant behavior data from a representative building to simulate behavior in a target building. A multi-dimensional histogram is constructed and populated with the recorded data. To handle sparse data, a smoothing technique is applied, where smoothing parameters are determined for influencing factors, and coefficients are computed for histogram bins based on these parameters and neighboring bins. The smoothed data is then normalized into arrays of occupant behavior features. Using these normalized arrays, the system generates attributes representing occupant behavior, such as movement patterns or device usage. These attributes are used to simulate occupant behavior in the target building, producing a prediction of energy demand. The simulation accounts for variations in behavior influenced by factors like time of day or environmental conditions, improving the accuracy of energy consumption forecasts. This approach enables more precise energy management and optimization in buildings.
11. The non-transitory computer-readable storage medium of claim 10 , wherein the one or more attributes include a task performed by the simulated occupant of the target building structure.
The invention relates to a system for simulating occupant behavior in a building structure to optimize energy efficiency. The problem addressed is the lack of accurate modeling of human behavior in buildings, which leads to inefficient energy use. The system simulates occupants performing tasks such as moving between rooms, adjusting lighting, or operating appliances, and uses this data to optimize energy consumption. The simulation includes attributes like task performance, which helps predict how occupants interact with the building's systems. By analyzing these simulated behaviors, the system can adjust HVAC, lighting, and other energy-consuming systems to reduce waste. The simulation accounts for variations in occupant routines, preferences, and environmental conditions to provide realistic energy-saving recommendations. This approach improves upon traditional static energy models by dynamically adapting to simulated human activity, leading to more efficient building operations. The system may also integrate real-time data to refine simulations and further enhance energy optimization.
12. The non-transitory computer-readable storage medium of claim 10 , wherein the one or more attributes include a time duration during which a task is performed by the simulated occupant of the target building structure.
This invention relates to building simulation systems that model occupant behavior within a target building structure. The problem addressed is the lack of realistic simulation of how occupants interact with a building over time, which is critical for energy efficiency, safety, and comfort analysis. The system simulates occupants performing tasks within the building, where each task is defined by one or more attributes. These attributes include a time duration during which the task is performed, allowing for realistic modeling of how long an occupant spends on activities such as using appliances, moving between rooms, or operating building systems. The simulation system generates a virtual representation of the building and its occupants, tracking their actions and interactions with the environment. The attributes may also include other factors like task frequency, location, and dependencies between tasks. By incorporating time duration as an attribute, the system improves the accuracy of simulations, enabling better predictions of energy consumption, occupant comfort, and building performance. This approach is particularly useful for architects, engineers, and building managers who need to optimize building designs and operations based on realistic occupant behavior patterns.
13. The non-transitory computer-readable storage medium of claim 10 , wherein the one or more attributes include a number of participants associated with a simulated task performed by the simulated occupant of the target building structure.
This invention relates to a computer-readable storage medium for simulating occupant behavior in a target building structure, particularly for evaluating building performance. The system simulates occupants performing tasks within the building, such as moving between rooms or operating equipment, to assess factors like energy usage, safety, and efficiency. A key aspect is the ability to model various attributes of these simulated occupants, including their interactions with the building environment. One specific attribute addressed is the number of participants involved in a simulated task. This allows the system to analyze how different group sizes impact building performance. For example, simulating multiple occupants performing a task simultaneously can reveal bottlenecks in space utilization or energy consumption patterns. The system can adjust this attribute to test scenarios with varying numbers of participants, providing insights into optimal occupancy levels for different building designs or operational strategies. The simulation may also incorporate other attributes, such as occupant movement patterns, task completion times, or interactions with building systems, to create a comprehensive evaluation of the building's performance under realistic conditions. This approach helps architects, engineers, and facility managers optimize building designs and operations based on simulated human behavior.
14. The non-transitory computer-readable storage medium of claim 10 , wherein generating one or more attributes comprises: computing a probability table based on the normalized arrays of occupant behavior feature values; computing one or more gamma distribution parameters based on the normalized arrays of occupant behavior feature values; and generating a first occupant behavior attribute for a sub interval based on the probability table or the gamma distribution parameters.
This invention relates to analyzing occupant behavior in a building or space using sensor data. The system collects sensor data from various sources, such as motion sensors, environmental sensors, or other monitoring devices, to detect and track occupant activities. The challenge addressed is accurately identifying and quantifying occupant behavior patterns to optimize energy efficiency, security, or space utilization. The system processes raw sensor data to extract behavior feature values, which are then normalized into arrays. These normalized arrays are used to compute a probability table and gamma distribution parameters, which model the statistical distribution of occupant behaviors. The system generates occupant behavior attributes for specific time intervals by analyzing the probability table or gamma distribution parameters. These attributes may include activity frequency, duration, or other behavioral metrics. By leveraging probabilistic and statistical modeling, the system provides insights into how occupants interact with the space, enabling applications such as automated lighting control, HVAC optimization, or security monitoring. The approach improves upon traditional methods by incorporating dynamic, data-driven behavior modeling rather than static rules.
15. The non-transitory computer-readable storage medium of claim 10 , wherein a number of dimensions of the histogram equals a number of factors that influence the one or more attributes.
This invention relates to data analysis systems that use histograms to model and analyze complex datasets. The problem addressed is the difficulty in accurately representing multi-dimensional data distributions, particularly when multiple factors influence the attributes being analyzed. Traditional histograms often fail to capture the full complexity of such datasets, leading to incomplete or misleading insights. The invention improves upon prior art by generating a histogram with a number of dimensions equal to the number of factors influencing the analyzed attributes. This ensures that the histogram accurately reflects the multi-dimensional nature of the data. The system first identifies the relevant factors affecting the attributes, then constructs a histogram with corresponding dimensions. This allows for precise modeling of how different factors interact to influence the attributes. The histogram can then be used for further analysis, such as identifying patterns, outliers, or correlations within the data. The invention also includes methods for dynamically adjusting the histogram dimensions based on changes in the dataset or the factors being analyzed. This ensures the histogram remains accurate as new data is introduced or as the relationships between factors evolve. The system may also include visualization tools to display the multi-dimensional histogram in a user-friendly format, aiding in data interpretation. The overall approach provides a more accurate and flexible way to analyze complex datasets where multiple factors influence the attributes of interest.
16. The non-transitory computer-readable storage medium of claim 15 , wherein the factors comprise at least one of time of day, the previous task performed by an actual occupant, time elapsed since a task was performed by the actual occupant, and a number of occupants participating in a task.
This invention relates to a system for optimizing task performance in a smart environment by analyzing occupant behavior and environmental conditions. The system uses a computer-readable storage medium to store instructions for a processor to determine task performance factors, such as time of day, the previous task performed by an occupant, the time elapsed since a task was performed, and the number of occupants participating in a task. These factors are used to predict the likelihood of a task being performed by an actual occupant versus an automated system. The system then selects an appropriate action based on this prediction, such as adjusting environmental controls, scheduling tasks, or prompting the occupant. The goal is to improve efficiency and user experience by dynamically adapting to occupant behavior and preferences. The system may also learn from historical data to refine its predictions over time. This approach reduces unnecessary automation while ensuring tasks are completed efficiently.
17. A computing system configured to simulate occupant behavior for a target building structure, comprising: a memory configured to store an application program and recorded occupant behavior data associated with a representative building structure associated with the application program; and a processor that is configured to: receive the recorded occupant behavior data associated with the representative building structure, construct a histogram of at least two dimensions that is populated with the recorded occupant behavior data, smoothing the recorded occupant behavior data to fully populate the histogram, wherein the smoothing comprises: determining a smoothing parameter value for each of one or more factors that influences the one or more attributes; and computing a smoothing coefficient for each of the plurality of bins based on the smoothing parameter values and a number of neighboring bins, compute normalized arrays of occupant behavior feature values based on the histogram; generate one or more attributes of occupant behavior for a simulated occupant of the target building structure based on the normalized arrays of occupant behavior feature values, store the one or more attributes of occupant behavior in the memory, and generate a simulation of occupant behavior based on the generated one or more attributes of occupant behavior, wherein the simulation indicates a predicted energy demand of the target building structure.
This technical summary describes a computing system designed to simulate occupant behavior in a target building structure to predict energy demand. The system addresses the challenge of accurately modeling how occupants interact with building systems, which is critical for energy efficiency and demand forecasting. The system includes a memory storing an application program and recorded occupant behavior data from a representative building structure. A processor receives this data and constructs a multi-dimensional histogram populated with the recorded behavior data, applying smoothing techniques to ensure full population of the histogram. Smoothing involves determining parameter values for factors influencing occupant behavior attributes and computing smoothing coefficients for histogram bins based on these parameters and neighboring bins. The system then computes normalized arrays of occupant behavior feature values from the histogram. Using these normalized values, the system generates attributes of occupant behavior for a simulated occupant in the target building structure, stores these attributes, and produces a simulation of occupant behavior. The simulation outputs a predicted energy demand for the target building structure, enabling better energy management and optimization. The system leverages historical behavior data and statistical smoothing to improve the accuracy of energy demand predictions.
18. The computing system of claim 17 , wherein a number of dimensions of the histogram equals a number of factors that influence the one or more attributes.
A computing system is designed to analyze and visualize data attributes influenced by multiple factors. The system generates a histogram with a number of dimensions equal to the number of influencing factors, allowing for multi-dimensional data representation. This approach enables users to explore complex relationships between attributes and their influencing factors in a structured manner. The histogram can be dynamically adjusted to reflect changes in the data or user-defined parameters, providing real-time insights. The system also includes a user interface that allows users to interact with the histogram, such as filtering, zooming, or selecting specific data points for further analysis. The histogram can be displayed in two or three dimensions, depending on the complexity of the data and user preferences. The system is particularly useful in fields like data science, analytics, and decision-making, where understanding the interplay between multiple factors and their impact on attributes is critical. The multi-dimensional histogram helps users identify patterns, trends, and outliers that may not be apparent in traditional two-dimensional representations. The system can be integrated into existing data analysis tools or deployed as a standalone application.
Unknown
March 10, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.